o A ■ STUDIES IN INFLUENZA AND PNEUMONÍA IV. Further Results of Prophylactic Inoculations •¿J <* L jUHAKv E. C. ROSEN O W, M.D. AND B. F. S T URDÍ V A N T. M.D. ROCHESTER, MINX. STUDIES IN INFLUENZA AND PNEU- MONÍA IV. FURTHER RESULTS OF PROPHYLACTIC INOCULATIONS * E. C. ROSENOW, M.D. AND B. F. STURDIVANT, M.D. ROCHESTER, MINN. To determine the valué of vaccination against dis- ease, it is essential that the disease shall be one which a relatively large number of persons will develop unless protected, and that it be accompanied by serious conse- quences. These conditions were amply fulfilled dur- ing the pandemic of influenza. Moreover, the vaccine should contain the killed bacteria that produce symp- toms and which are at least contributory to the cause of death. \Ye have attempted, so far as possible, to fulfil this requirement by making a careful bacteriologic study of the disease, and by incorporating into the vac- cine the important bacteria isolated. The epidemic was severe, and the need and the demand for vaccina- tion were great; a large number of cases were available for bacteriologic study and to supply the proper strains for the vaccine. Vaccinations in large numbers during the past ten years with bacteria belonging to the group found in influenza have at least proved harmless, and in the case of pneumonía, prophylactic vaccinations * From the División of Experimental Bacteriology, Mayo Foun- dation. * This paper and that of Dr. G. W. McCoy which follows are part of a symposium on "Influenza." The remaining papers and the discussion will appear next week. * Read before the joint meeting of the Section on Pharmacology and Therapeutics, the Section on Pathology and Physiology and the Section on Preventive Medicine and Public Health at the Seventieth Annual Session of the American Medical Association, Atlantic City, N. J., June, 1919. 4 have been successfully carried out by Wright,1 Lister,2 and Cecil and Austin.3 A splendid opportunity to study the effect of prophylactic moculation was at hand. Owing to the foresight of the founders of the Mayo Foundation, necessary funds to meet the emer- gency were available. A large amount of the vaccine has been prepared and sent gratis on request to numer- ous physicians on condition that reports of the results be returned. In a previous report,4 the reasons for the use of a mixed vaccine containing, as far as possible, freshly isolated strains were discussed. It was pointed out that the streptococci, especially green-producing strep- tococci from influenza, have certain peculiar proper- ties. The preliminary results, as reported from the use of this vaccine, indícate that considerable protection is afforded against influenza and especially against the accompanying pneumonía. Vaccinations were begun soon after the onset of the epidemic. The period" of observation was six weeks. It is our purpose in this paper to emphasize essential points in the preparation of the vaccine, to present further results from its use, and to record certain immunologic experiments. COMPOSITION AND PREPARATION OF THE VACCINE Influenza bacilli were isolated in large numbers at the outset of the epidemic, but they were rarely found later in the epidemic. The small fraction of influenza bacilli included in the first few batches of vaccine were therefore omitted, and the vaccine was made to con- tain a proportionately higher percentage of the green- producing streptococci. In other respects, the original formula has been adhered to. The formula as used in almost all cases covered by the present report is given in Table 1. 1. Wright, A. E.; Morgan, W. P., et al.: Observations on Prophy- lactic Inoculation Against Pneumococcus Infections and on the Results Which Have Been Achieved by It, Lancet 1: 1-10 (Jan. 3) 1914. 2. Lister, F. S.: Prophylactic Inoculation of Man Against Pneumo- coccal Infections and More Particularly Against Lobar Pneumonía; Including a_ Report on the Results of the Experimental Inoculation, with a Specific Group Vaccine, of the Native Mine Laborers Employed on the Premier (Diamond) Mine and the Crown (Gold) Mines in the Transvaal and the de Beers (Diamond) Mines at Kimberley—Covering the Period from Nov. 1, 1916, to Oct. 31, 1917, Publications of the South African Institute for Medical Research, Johannesburg, South África, W. E. Hortor and Company, Ltd., 1917, pp. 1-30. 3. Cecil, R. L., and Austin, J. H.: Prophylactic Inoculation Against Pneumococcus, J. Exper. M. 38:19-41 (July 18) 1918. 4. Rosenow, E. C.: Prophylactic Inoculation Against Respiratory Infections: Preliminary Report, J. A. M. A. 72:31-34 (Jan. 4) 1919. 5 The preparation of the médium, the method of culti- vating and collecting the bacteria, and the procedure of standardizing the dose and killing the bacteria are described in the preliminary report.4 The vaccine, it will be remembered, was made to contain approxi- mately 5 billion bacteria for 1 ce. Later, the concen- traron was made twice as great, and the quantity of liquid was reduced to one-half. The injections were given subcutaneously one week apart. The first dose of the concentrated vaccine (0.25 ce.) contained 2.5 billion, the second (0.5 ce) 5 billion, and the third (0.75 ce) 7.5 billion bacteria. Considering the large size of these doses and the reactions obtained, the injections should not be given of tener than once a week in order not to overstimulate the mechanism of immunity. TABLE 1— FORMULA OF VACCINE Pneumococci, Types I (10 per cent.), II (14 per cent.), and III (6 per cent.)........................................... 30 per cent. Pneumococci Group IV and the allied green-producing diplostreptococci described .............................. 40 per cent. Hemolytic streptococci ....................................... 20 per cent. Staphylococcus aureus ....................................... 10 per cent. The tendeney of streptococci to undergo changes and to lose specific properties has been repeatedly emphasized by one of us. It was thought important that freshly isolated strains should be included in the vaccine. In Table 2 are given the culture generations of all the strains that have been used throughout the epidemic. The fermentation power was tested of fifty-seven strains of the green-producing streptococci included in the vaccine; only twenty-seven fermented inulin, and only eight were bile soluble. The advantages which should come from the use of a lipovaccine, particularly when a series of strains needs to be included, have already been pointed out, and a simple method for the preparation in oil of a vaccine of the formula given in Table 1 has been developed and submitted for publication. A further study of the sputum and other material shows that of all the bacteria isolated, the somewhat peculiar green- producing streptocoecus or diplostreptococcus is the most important. This organism is present in large numbers at the very outset of symptoms of influenza and of the accompanying pneumonía; it is commonly 6 present after death. If the sputum or mass cultures are injected íntraperítoneally into animáis, they die, usually from invasión of the green-producing strepto- cocci or pneumococci. If injected intratracheally in guinea-pigs the picture of influenzal pneumonía is closely simulated. Immunologic experiments with the serum from a horse injected with one strain indícate that most of the strains are immunologically alike. The serum of cases of influenza develops agglutinating power over these strains. AGGLUTINATING POWER In Table 3 it is shown that the vaccine used pos- sessed well marked antigenic powers. The strains S 1, S 3, 25982.2, 2604.2, 3048.3, and 28742.3 were green- producing streptococci or pneumococci; 2575.2, a hemolytic streptococcus, and 26083.2, a staphylococcus from cases of influenza. It will be noted that agglu- tinins appear in the serum on the tenth day and persist for six weeks. Table 3 shows, moreover, that TABLE 2.—CULTURE GENERATION OF BACTERIA FROM INFLUENZA AS USED IN THE VACCINE Cultures Green-Pro-ducing Strep-tococcus Hemolytic Strepto-coccus Staphylo-coccus Eleventh to twentieth generation 58 95 21 18 20 0 18 8 0 Total......................... 174 38 26 the bacteria in the vaccine (492) used as the antigen in the first column were susceptible to agglutination. This vaccine was prepared three months previously and was kept in the ice chest. Most of the strains used as antigen in the experiment recorded in this table were not included in the vaccine used to immunize the persons whose serums were tested. All the green- producing streptococci were agglutinated, however, by the monovalent horse serum. In Table 4 are given the results following the injec- tion of a single dose of the lipovaccine (from 25 to 75 billions) in three persons. It may be noted that the amount of agglutination is greater than that following the injection of the saline vaccine, but here, as in the TABLK 3.-AGG LUTINATING l'OWHR OF THK SKRUM OF l'KRKONS TNOCULATKD WITH SALINK VACCINl -Strains- Serums (Dilutions 1:20) 492 SI S 3 2,5982.2 2,004.2 3,048.3 2,8742.3 2,575.2 2,608^.2 2",542 24 hours before third dose saline vaccine.............................. 0 0 00 0 + + 0 0 2,542 24 hours after third dose saline vaccine................................ 00000 + + 0" 2,542 48 hours after third dose saline vaccine................................ 0 0 0 0 4-4- ++ + + 0 v 2,542 10 days after third dose saline vaccine................................. + ++ + 0 ++ ++ + + + ++ u 2,542 6 weeks after third dose saline vaccine............................... + ++ •■ •• + + •■ •• •• u 2,543 24 hours before third dose saline vaccine.............................. 0000 + 0000 2,543 24 hours after third dose saline vaccine................................ 00004-000 2,543 48 hours after third dose saline vaccine................................ + 000 +4- 0 +4- 0 0 2,543 10 days after third dose saline vaccine................................. 4-4- +00 ++ ++ +++ 4- + 2,543 «weeks after third dose saline vaccine............................... ++ +00 +++ ++ 4-4- 4- 0 2,545 24 hours before third dose saline vaccine.............................. 00 0 000000 2,545 24 hours after third dose saline vaccine................................ 0000000 0 2,545 48 hours after third dose saline vaccine................................ 0000 + 0000 2,545 10 days after third dose saline vaccine................................. ++ 0 0 0 ++ ++ ++ ++ + + 2,545 6 weeks after third dose saline vaccine............................... ++ 0 .0 + ++ ++ ++ ++ + + 2,547 24 hours before third dose saline vaccine.............................. 000000000 2,547 24 hours after third dose saline vaccine................................ 000000000 2,547 48 hours after third dose saline vaccine................................ + 0 0 0 + + + + + 0 0 2,547 10 days after third dose saline vaccine................................. ++ 0 + + ++ ++ ++ 0 0 2,547 6 weeks after third dose saline vaccine............................... ++ 0 + + ++ ++ ++ 0 ++ 3,075 normal ................................................................ 0 0 0 0 0 0 0 0 0 3,076 normal ................................................................. 0 0 0 0 0 0 0 0 0 NaCl .......................................................... 0 0 0 0 0 0 0 0 0 case following the injection of the saline vaccine, not all strains are equally susceptible to agglutination, and some are not agglutinated at all. TABLE 4.—AGGLUTINATING POWER OF THE SERUM OF PER- SONS INOCULATED WITH LIPOVACCINE Serums ,-----------------Strains-----------------> (Dilutions 1:20) 3,2712.3 3,2962.2 3,331 3,332.2 3,334.2 3,334.2 3,342 3,074 normal.............. 0 ++ + + 0 ++ 0 3,074 4 days after lipo- vaccine.................. + + + + ++ ++ + ++ 0 3,074 10 days after lipo- vaccine.................. + + + + + + + +++ ++ ++ + + + 0 3,074 6 weeks after lipo- vaccine.................. ++ + + + ++ + + ++ 0 3,075 normal.............. 0 ++ 0 0 + 0 0 3,075 4 days after lipo- vaccine.................. ++ +00+ ++ 0 3,075 10 days after lipo- vaccine.................. + + + + + + + ++ + + 0 3,076 normal.............. 0 ++ + 0 0 0 0 3,076 4 days after lipo- vaccine.................. + + + + +0 + 00 3,076 10 days after lipo- vaccine.................. ++ +++ ++ 0 + 0 0 3,076 6 weeks after lipo- vaccine.................. + + + + + + + + 0 ++ 0 0 NaCl...................... 0 0 0 0 0 0 0 Table 5 shows the agglutinating power of various immune horse serums over strains of green-producing streptococci from influenza, strains included in the vac- cine. The serum from Horse 15, immunized with one strain from the blood of a patient who died, has marked agglutinating power over most of the strains. Of the thirty-three strains tested in this manner, twenty-five were agglutinated specifically by this serum. The results indícate clearly that among the green-pro- ducing streptococci, including Group IV pneumococci in influenza, there are strains which have a specific relationship, and that we were fortúnate in success- fully separating them from the ordinary Streptococcus viridans and including them in the vaccine long before the results of immunologic experiments were available. The apparent protection against attacks of influenza noted in the preliminary report, difficult to understand at that time, now becomes rational. METHOD OF SECURING DATA In most instances the reactions were mild, about one person in each 100 reacted more severely. Some reacted severely to all three inoculations, others only AGGLUTINATING POVVKR OF VARIOUS IMMUNE HORSE SKRUMS OVKR STREPTOCOCCI INCLUDED IN THE VACCINE Serums ,------------------------------------ (Dilutions 1:20) 2,347.19 2,349.13 2,350.16 2,531.14 2,532.4 2,534.11 Pneumococcus Type I— 0 0 0 + 00 Pneumococcus Type II... ++ 0 0 + 0 + Pneuniococcus Type 111.. 0 0 0 + 0 + Horse 9................... ++ ++ 0 + 0 + + Horselü..................+ + + + + + + +4- + + + + + + + + + + + + + Normal horse............. 0 0 0 + 0 + NaCl...................... 0 0 0 0 0 0 -Strains- 2,001.2 0 0 0 0 + + + 0 o o o o o ++++ o o o ++ o o ++ o o 2,(¡I82.2 2,684.16 2,698-,3 0 0 o o ++++ o o 2,719a 0 0 o +++ ++ o o 2,769 0 0 o o o o o ,800-. o o o o ++ o o 2,825 0 0 0 + + + + + 0 o TABLK 6.-RKSULTS AS RKPORTED IN QUENTIONNAIRKS FROM ALL SOURCES Groups Vaccinated once... Vaccinated twice.. Vaccinated 3 times Not vaccinated— Total Numbi'i 26,936 23,348 93,476 345,133 Influ- enza 118.2 97.0 87.9 281.8 -Disease- Acute Edema Pneu- of Lungs monia 3.1 0.77 0.8 4.1 8.7 3.04 4.4 21.0 —Incidence for 1,000 Persons- Empy- ema 0.29 0.17 0.18 0.83 Acute Edema of Lungs 0.14 0.47 0.18 1.7 Pneu- monía 2.6 1.9 1.2 2.37 -Deaths- Empy- ema 0.07 0.04 0 0.07 Menin- gitis 0.18 0.21 0.05 0.15 Encepha- litis 0.03 Total Deaths 3.0 2.62 1.43 8.55 10 to one or two. Persons coming down with a cold or with symptoms of influenza are often hypersensitive. Marked diffuse redness resembling erysipelas about the site of inoculation, with swelling and, later, marked induration, has occurred occasionally. In no instance were the symptoms alarming. The number of severe reactions is sufficiently large, however, to prevent gen- eral vaccination except at the time of an acute emer- gency. This is in accord with the experience of Cecil and Austin,3 noted during prophylactic inoculations with pneumococci. An outline for records of persons vaccinated was sent with each batch of vaccine and later a questionnaire. The questionnaire asked for the date of the onset of the epidemic, the date when the vaccine was first used, the week of the height of the epidemic, the week in which the greatest number of vaccinations were given, and the duration of the epi- demic. The number of cases of influenza from the time the vaccinations were begun until the end of the epidemic, or up to May 1, and the number of deaths which occurred among the vaccinated and unvaccinated in the same period, in the practices of the physicians supplied with the vaccine, were asked. The reports of the use of the vaccine after the epidemic had disap- peared were excluded. The period of observation in most instances was from four to five months. In determining a safe criterion as to the valué of the vaccine, we have purposely been unfair to the vaccin- ated group. The protection afforded among the vac- cinated patients was measured from the day of the first vaccination, whereas, judging by the agglutination experiments, it should be calculated from about one week after the third injection. There is another reason why we have arbitrarily decided to make our calculations from the day of the first vaccination. A procedure, calculated to protect against an epidemic disease, such as influenza, should have sufficient protective valué when given after the onset of the epidemic to be measurable, for it is prac- tically impossible to anticípate these epidemics and, moreover, persons will not present themselves for vac- cination Until the epidemic is at hand. The questionnaire was arranged so as to yield infor- mation regarding the incidence of influenza, acute. • 11 edema of the lungs, pneumonía and empyema, and the deaths from acute edema of the lungs, pneumonía, empyema, meningitis, and encephalitis among the vac- cinated and the unvaccinated. Sepárate reports includ- ing the foregoing points were asked for from institu- tions and in the cases of pregnant women. The impressions gained from the use of the vaccine regard- ing the severity of the disease if contracted following vaccination, and the effect, if any, which the vaccine had on certain chronic infections. such as bronchitis, sinusitis, myositis, and arthritis were asked for. Many physicians were so overwhelmed during the height of the epidemic that accurate records could not be kept, and accordingly the reports containing accu- rate data are proportionately few. The reports of 530 physicians were fairly complete, however, and these are summarized in Table 6. It is realized that there must necessarily be errors in the morbidity figures as reported to us, just as in the case of reports to boards of health. It is generally agreed that as influenza became more prevalent and less severe, a proportion- ately smaller number of cases were reported, and that all morbidity figures reported are well below the actual figures. The error, however, among the vaccinated and unvaccinated groups in the reports to us, should be approximately the same, and henee the figures should be comparable. Mortality figures, on the other hand, may be considered as fairly accurate. RESULTS OF INOCULATION The total number of unvaccinated persons recorded in Table 6 represents the sum of the estimated dién- teles of the various physicians reporting the cases, and averages about 1,200 for each. It will be noted that the incidence of influenza, of acute edema of the lungs, of pneumonía following influenza, and the number of deaths from all causes among the vaccinated are con- sistently lower than that among the unvaccinated. Moreover, the incidence of disease and deaths is lowest in the group of 93,476 persons who were vaccinated three times. The reports included in this table were from many states, but the largest number carne from Iowa, Minnesota, and Wisconsin. Thirteen thousand, six hundred and fifty persons inoculated and 2,083 who died were grouped according to age by decades. The 12* curves indicating the percentage in each run roughly parallel. The largest number of inoculations were given and the largest number of deaths occurred between the ages of 11 and 40 years. The percentages of the for- mer in these three decades were 23, 19, and 21, of the latter 13, 29, and 23, respectively. Through the coop- eration of the Board of Health of Minnesota we were able to check the results as reported to us with the morbidity and mor^lity figures as reported to them. Reports on a considerable number of vaccinations were received from Brown, Chippewa, Clay, Dodge, Fill- more, Goodhue, Houston, Itasca, Lesueur, Lyon, Mower, Olmsted, Rice, Stearne, Steele, Wabasha, Waseca, Watonwan, and Winona counties. The total estimated population of these counties is 472,584. The total number of cases of influenza in these counties reported to the board of health from the beginning of the epidemic until May 1 is 30,763, or sixty-five for each thousand. This is admittedly a low figure. The total mortality rate as reported to the board of health during this time is 4.2. The mortality rate, excluding the deaths which occurred in the respective counties prior to the date of the first vaccinations, is 3.2 (Table 7). The figures in the table indicating the cases and the deaths as reported to us are believed to be more accu- rate. The mortality rate, exclusive of that of the Mayo Clinic, in the 17,532 persons vaccinated three times is only one fourth of that reported to the board of health. Moreover, the total number of deaths among the vaccinated, including the per- sons inoculated only once and twice, is 1.6 for 1,000, or half the mortality rate as reported to the board of health during the same period of time. When we consider the fact that the deaths in each group were counted from the time the first vac- cinations were given, which is really unfair to the vaccine, and the fact that our figures include all pneu- monías, while those of the board of health include only the influenzal pneumonías, there seems little doubt that the difference must be due to the protection afforded by the vaccine. The figures given in Table 8 for Olm- sted County, where about one third of the population was vaccinated, exclusive of the Mayo Clinic and the state hospital, are similar to those obtained elsewhere. TABLE 7. -RESULTS AS REPORTED IN QUESTIONNAIRES FROM NINETEEN COUNTIES IN MINNESOTA EXCLUSIVE OF THE MAYO CLINIC Groups Vaccinated once... Vaccinated twice.. Vaccinated 3 times Not vaccinated.... As reported to State Board of Health........... 472,584 Total Number 4,828 4,02